XCOM 2
Second Wave Reborn
Statistical Analysis on NCE game balance & my own take at balancing the mod
Hey everyone. I loved Not created equally in EW and in fact, I actually didn't buy XCom2 until this mod got released (i followed EladDv status updates over at the nexusmods boards). Not only is this mod great and works flawlessly, you can even edit it to your liking. Just open the XComSecondWave.ini file in your \steamapps\workshop\content\268500\622475105\Config folder, and edit away! But be careful, balancing these values is actually harder than it looks (and probably needs extensive testing or simulations if you are no math genius). It's very easy to create a .ini, that leads to unbalanced results.

So why did i create this discussion? I wrote a small amateurish javascript code snippet that lets me simulate, what stats soldiers get with any given ini. Here are some of my findings, after generating 10000 soldiers with the default values of the mod:

  • about 40% of all rookies have above 64 aim
  • about 4% of all rookies have aim>69 & mobility>12 (these are usually the most desired rookies)
  • about 15% of all rookies have aim<60 & mobility<12 (these suck, although they may have other desirable stats like high hp etc.)

So on the first glance, the mod actually makes the game harder, since there are more rookies with below average aim, than ones with above average aim. Additionally there are many really bad ones but comparatively few rookies with well rolled stats. But one has to consider, that rookies can now roll stats, they dont have acces to in Vanilla Xcom 2. These stats are: higher flanking crit chance, Defense>0, Dodge>0 (min dodge is actually 5 in the default settings), and armor mitigation>0. All in all the on average worse aim of rookies is probably counterbalanced by these stats.

However I myself prefer a balancing that is nearer to XCom:EW Vanilla NCE. After countless simulations and playtesting i came up with my own version of the .ini file that i now feel confident enough to publish. My main goals were:

  • create more variance between rookies to create a style of NCE that is nearer to the XCOM:EW experience without being imbalanced (EW tends to be easier with NCE activated, because rookies are cheap so you can screen extensively)
  • have slightly more "really good" rookies and less "really bad ones"
  • diminish the effect of newly introduced stats that are potentially gamebreaking, especially in the lategame (dodge, armor mitigation!, flanking crits)
  • i also am not fond of variable stats, that are not displayed ingame. I want to make educated tactical decisions, but I cannot do so if I dont know, that my rookie has for example higher than normal flanking crit chance

Therefore i removed armor mitigation (hard to balance) and higher than normal flanking crit chance. I also toned down defense and dodge. I do find the occasional lucky grazed shot exciting, but i dont want it to occur too often. I reintroduced the stat ranges of EW for the following stats: mobility, Offense, Will. There are some more smaller changes which you can read below. Here are my values:

HP_HIGH=5
HP_LOW=4
HP_COST=6
Offense_HIGH=80
Offense_LOW=50
Offense_COST=2
Defense_HIGH=4
Defense_LOW=0
Defense_COST=2
Mobility_HIGH=14
Mobility_LOW=10
Mobility_COST=7
Will_HIGH=55
Will_LOW=25
Will_COST=1
Hacking_HIGH=10
Hacking_LOW=0
Hacking_COST=0
Dodge_HIGH=4
Dodge_LOW=0
Dodge_COST=2
ArmorMitigation_HIGH=0
ArmorMitigation_LOW=0
ArmorMitigation_COST=15
SightRadius_high=27
SightRadius_low=27
SightRadius_Cost=0
FlankCrit_high=40
FlankCrit_low=40
FlankCrit_Cost=0

TOTALPOINTS=66 ; Total Points allocated to stats.
TOLERANCE=24

Some stats from simulations i did with these values:
  • about 47% of all rookies have above 64 aim
  • about 4.6% of all rookies have aim>69 & mobility>12 (these are usually the most desired rookies)
  • about 7.7% of all rookies have aim<60 & mobility<12 (these suck)

The mod is balanced for commander and legendary difficulty. Lower difficulties would need slight alterations because rookies have higher HP.

As you can see these value are very balanced. Rookies are on average still slightly worse than vanilla Rookies, but they do get some other small benefits to compensate (dodge, Defense, possibly higher hp). I do have to say, that NCE in general works even better in XCOM 2, than in EW, because especially in the early stages of the game, you eventually have to use all your soldiers. And its great, when you actually make a Sniper with 52 aim and 12 mobility work (focussed this one on pistols). This may not hold true for difficulties below legendary, but i have no experience with these.

Wow this was longer than i expected it to be. Feel free to use my values, if you want. But keep in mind: as long as you are subscribed to the mod, everytime it updates, the ini resets to its default values. So save the values and check the ini every time before you start the game. Also, if you did some ini editing yourself i would like to read your take on balancing the values.
Last edited by LeaderEnemyBoss; 1 Mar, 2016 @ 4:05pm
< >
Showing 1-15 of 19 comments
SShaker 18 Feb, 2016 @ 4:02pm 
Ok I want to use your values because they seem balanced, I want to do a run on Legend with 6-8 squad size and more soldiers like Long War.

Now I changed the stats in the XComSecondWave.ini and made a copy of it with the modified stats. What will happen when this mod gets updated though? Is there a way to keep the settings, because all I am interesed in is the NCE option, I am using the other hidden potential mod.

Thanks for the testing and TY to the mod author!
LeaderEnemyBoss 19 Feb, 2016 @ 1:09am 
When the mod gets updated, the ini resets to the default values. EladDv seems to slow down on updating in favor of more thorough testing, so it isnt like in the beginning, where i had to constantly check for ini changes. But as long as you are subscribed to the mod, you should check the ini before each start of the game. As soon as you you notice the old values, just replace it with my settings again.

I dont know if it is possible to play with a mod, if you unsubscribe from the workshop. I tried copying the mod folder and than unsubscribed, but the mod wouldn show up in the XCOM 2 launcher. If you can make a mod work with unsubscribing, you can basically disable the autoupdate feature of the mod and wont have to look at the ini anymore.
LeaderEnemyBoss 19 Feb, 2016 @ 2:59am 
Originally posted by SShaker:
Ok I want to use your values because they seem balanced, I want to do a run on Legend with 6-8 squad size and more soldiers like Long War.

Now I changed the stats in the XComSecondWave.ini and made a copy of it with the modified stats. What will happen when this mod gets updated though? Is there a way to keep the settings, because all I am interesed in is the NCE option, I am using the other hidden potential mod.

Thanks for the testing and TY to the mod author!

Okay here is the solution (thx to EladDv) on how to use a manually installed mod. Copy the contents of the steam mod folder (\steamapps\workshop\content\268500\622475105) and put them in the following folder: steamapps\common\XCOM 2\XComGame\Mods\SecondWaveReborn\. Do not copy the folder itself, only its contents. Unsubscribe from the mode. You can now use the mod without being subscribed to it. Be aware, that you will not receive automatic mod updates. However you can ofcourse do a manual update from time to time (just subscribe, get the new files, put them in the Xcom 2 mod folder, change the ini values, and unsubscribe again).
Last edited by LeaderEnemyBoss; 19 Feb, 2016 @ 3:01am
GeminiGod 28 Feb, 2016 @ 3:08pm 
Interesting test you ran. Thanks for sharing. I agree it would be nice to have all the numbers visible for variable stats like with the Long War, but I recognize there are major programming hurdles there.
Last edited by GeminiGod; 28 Feb, 2016 @ 5:03pm
GeminiGod 28 Feb, 2016 @ 3:09pm 
On that note, Elad, I think it might be helpful for people debating what mod to download if you highlighted in the description that this mod is more in line with the Long War 2nd wave than EW, which is distinctly different from other mod options currently available.
Last edited by GeminiGod; 28 Feb, 2016 @ 5:04pm
GeminiGod 28 Feb, 2016 @ 6:59pm 
@Leaderenemyboss, I loved seeing your NCE stat choices and comparing with @EladDv. Figured I would share my own here for comparison. I am only just starting a campaign to see how it actually plays out, but these ranges and costs seem pretty balanced in my crazy logical mind.

I would prefer to use a negative defense low # like in LW, but until I am sure that this gets properly calculated for cost points and properly actuated in game I kept it at a simple 0-5. Also since there is no way presently to see the defense stat in the armory, it makes it too hard to properly appreciate soldiers with a high number (same goes for some of the other potentially modifiable stats).

I mostly balanced every stat around the vanilla number. One notable exception is mobility where I added a little more top-end opportunity given its importance with timer limts and carrying bodies (which I modded to have a -4 mobility penalty). Given that stun lancers have 15 mobility it only seems fitting my soldiers could too. This is still less than LW that capped mobility at 16. (For more info on mobility calcs: http://gaming.stackexchange.com/questions/256151/how-does-mobility-stat-correspond-to-tiles-moved-per-action).

A vanilla character with this setup would have 80 points. I upped total points to 90 because my rookies seemed to be coming up a little short on avg, especially given that aim is probably not properly weighted in the point distribution for its true value. I could compensate by upping the aim cap to 80 like @LeaderEnemyBoss did (and the way it is in EW), but I really feel like 80 is too high for a rooky in xcom2 (or xcom1 for that matter), especially with the powerful scope and PCS aim options in xcom2!

With the tolerance set at 30, this means that the best possible rookie is exactly twice as effective as the worst possible rookie. Seems about appropriate to RL IMO.

(EDIT: The figures below have been slightly modified since my original post based on the discussion that follows. (i.e. this will always show the most current version I am using.))

Here it is:

HP_HIGH=6 ; Vanilla: eStat_HP=5 (4 on commander & legend)
HP_LOW=3
HP_COST=10
Offense_HIGH=75 ; Vanilla: eStat_Offense=65
Offense_LOW=55
Offense_COST=2
Defense_HIGH=5 ; Vanilla: eStat_Defense=0
Defense_LOW=0
Defense_COST=2
Mobility_HIGH=15 ; Vanilla: eStat_Mobility=12
Mobility_LOW=10
Mobility_COST=10
Will_HIGH=55 ; Vanilla: eStat_Will=40 (I believe rookie is 50)
Will_LOW=25
Will_COST=1
Hacking_HIGH=20 ; Vanilla: eStat_Hacking=5
Hacking_LOW=0
Hacking_COST=1
Dodge_HIGH=5 ; Vanilla: eStat_Dodge=0
Dodge_LOW=0
Dodge_COST=2
ArmorMitigation_HIGH=1 ; Vanilla: eStat_ArmorMitigation=0
ArmorMitigation_LOW=0
ArmorMitigation_COST=20
SightRadius_high=27
SightRadius_low=27
SightRadius_Cost=0
FlankCrit_high=50 ; Vanilla: FlankCrit=40 except rookie is 50.
FlankCrit_low=30 ; For Absolutely Critical Change these two stats to 100
FlankCrit_Cost=1

TOTALPOINTS=90 ; Total Points allocated to stats.
TOLERANCE=30
Last edited by GeminiGod; 1 Mar, 2016 @ 1:29pm
LeaderEnemyBoss 29 Feb, 2016 @ 1:23am 
Out of curiosity i simulated your data, and they seem indeed to be very balanced, gj ;) !

about 44% of your rookies have average or higher aim, and about 5% are very good (aim >69, mob > 12). 10% of your rookies are less than desireable (aim<60, mob <12). Remember that vanilla hp in commander or legendary is actually 4. If you play on those levels, you probably may want to lower the hp maximum somewhat.


Concerning my upper aim limit: rookies with 80 or near 80 aim are exceedingly rare with my settings. I have no problem with the occasional super rookie, as long as there are some more sucky 52 aim 11 mobility guys to counterbalance everything ;).
GeminiGod 29 Feb, 2016 @ 1:47am 
Sweet! I was hoping you would simulate it. Thanks. Really helpful. I was a little concerned because I got lucky on a game I just started and had a bunch of high aimers.

RE: HP, yeah, I am going to change range to 3-6 instead of 3-7. I seem to recall LW being a range of 3-6 but could be wrong. Maybe that change will slightly boost my aim avg too.
LeaderEnemyBoss 29 Feb, 2016 @ 2:02am 
Always consider that your rookies can roll other goodies which they couldnt roll in vanilla (high hp and armor are huge benefits!). So a slightly lower average quality in rookies with respect to aim and mobility is actually good. With your proposed change, 47% of all rookies have average or higher aim, and 6% are really good, 7% are really bad. If more users ask for it, maybe i will make a user friendly version of my simulation code ;)
GeminiGod 29 Feb, 2016 @ 6:14pm 
Impressive dropping 1hp kicked up the avg that much. HP is an interesting stat. It is definitely very helpful in early game and is a nice tradeoff for bad aim. Survivability is pretty huge long term, though I would still definitely prefer a high aim with the time limits and all. Protracted fire fights are very problematic in xcom 2.
GeminiGod 29 Feb, 2016 @ 9:52pm 
I decided my flankcrit was too high for a rookie. I stupidly based the range around 50, but this is too high. Vanilla even on veteran is 40. FlankCrit_high=50 and FlankCrit_low=30 makes more sense IMO. This is more in line with the enemies who are flankcritting at 40 on commander and legend difficutly (33 on veteran difficulty).

That change shouldn't affect anything else in my script for total points or distribution/aim/mobility avg.
LeaderEnemyBoss 1 Mar, 2016 @ 2:25am 
Oh wow, i never knew that Flankcrit was difiiculty dependent. Just checked and can confirm. Gotta nerf my .ini then ;)
GeminiGod 1 Mar, 2016 @ 1:20pm 
I kicked up hacking high a tad to 20. This should weaken the soldiers ever so slightly while strengthing specialists whom I felt still were a little underpowered at end game. The punishment vs. reward of hacks were such that I rarely ever took a chance on hacking anything.

I'm not quite sure how hacking is calculated but it isn't a linear thing. 100 hacking doesn't mean you have a 100% chance of hacking anything unlike all the other stats in the game.
LeaderEnemyBoss 1 Mar, 2016 @ 4:04pm 
It gets calculated against the the hacking defense of the enemy. However i found if you train your hacker early, take that 20 hacking bonus as soon as you can and equip a skulljack, shutting down enemies becomes pretty reliable.
Last edited by LeaderEnemyBoss; 1 Mar, 2016 @ 4:05pm
LeaderEnemyBoss 5 Mar, 2016 @ 9:53am 
I startet a new playthrough with a slightly modified version of my old config

HP_HIGH=5
HP_LOW=3
HP_COST=6
Offense_HIGH=80
Offense_LOW=50
Offense_COST=2
Defense_HIGH=0
Defense_LOW=0
Defense_COST=2
Mobility_HIGH=14
Mobility_LOW=10
Mobility_COST=7
Will_HIGH=55
Will_LOW=25
Will_COST=1
Hacking_HIGH=15
Hacking_LOW=0
Hacking_COST=0
Dodge_HIGH=5
Dodge_LOW=0
Dodge_COST=2
ArmorMitigation_HIGH=0
ArmorMitigation_LOW=0
ArmorMitigation_COST=15
SightRadius_high=27
SightRadius_low=27
SightRadius_Cost=0
FlankCrit_high=40
FlankCrit_low=40
FlankCrit_Cost=0

TOTALPOINTS=67 ; Total Points allocated to stats. Lower values make the average rookie worse (be careful!)
TOLERANCE=22

I removed the variation in defense (since you cant check it ingame the variation in this stat is pretty uninteresting) and made hp rolls of 3 possible. This makes the early game a bit harder, since 3 hp rookies are almost garantueed onehits, but I like it so far, since those squishies tend to have good aim/mobility stats.

Other smaller changes:
-Hack can roll up to 15, to make it a bit more relevant (its very rare that this makes a difference, allthough im still waiting for that odd ranger,that gets 3 lucky hack increases and becomes the frst non-specialist hacker ;) )

-Dodge_HIGH up to 5 to account for loss of Defense, also i like rare grazed shots ;).

-small modifications to TOTALPOINTS and TOLERANCE to account for higher vanilla rookie cost.
< >
Showing 1-15 of 19 comments
Per page: 1530 50